This paper considers the problem of designing static-state feedback la
ws for output regulation of square affine nonlinear systems. The appro
ach taken is to use input-output decoupling techniques to simplify the
output regulation task into separate single-input single-output regul
ation tasks. In the case where the input-output decoupling matrix is f
ull-rank, this approach yields the well-known input-output linearizing
feedback law. In the case where the input-output decoupling matrix is
rank-degenerate, it is shown that a static-state control law for outp
ut regulation can be constructed as long as the system can be input-ou
tput decoupled via dynamic feedback. The internal stability of the clo
sed-loop system obtained using this approach is analysed. (C) 1997 Els
evier Science Ltd.